Smoking , atopy , and laboratory animal allergy KATHERINE

This study examined data from three cross sectional surveys of 296 laboratory workers exposed to small mammals. Four indices of laboratory animal allergy were studied: symptoms suggestive of occupational asthma, symptoms suggestive of any occupational allergy, skin weals to animal urine extracts, and serum binding in radioallergosorbent tests with urine extracts. Pooled data from the three surveys showed an association between smoking and all indices except radioallergosorbent tests; the association was significant for symptoms of occupational asthma. One of the three surveys consistently showed a stronger association of allergy indices with smoking than with atopy (defined on skin tests with non-animal aeroallergens). Associations with smoking persisted after stratifying by atopic status, suggesting that smoking may be a risk factor for laboratory animal allergy. Smoking is associated with asthma caused by enzyme detergents' and snow crabs,2 with skin weals to Bacillus subtilis? and coffee bean,4 and with serum IgE antibody against ispaghula4 and tetrachlorophthalic anhydride (TCPA).5 This suggests that smoking is a risk factor for occupational allergy, possibly by increasing mucosal transport of antigen.56 The study of TCPA workers showed a statistically significant interaction between smoking and atopy (defined on skin tests with common aeroallergens): 16% of 31 atopic smokers had IgE antibody against TCPA, 12% of 111 non-atopic smokers, 8% of 36 atopic nonsmokers but none of 98 non-atopic non-smokers.5 In animal experiments inhalation of tobacco smoke and other bronchial irritants potentiates sensitisation by inhaled antigen.'-9 Laboratory animal allergy (LAA) is common in research workers, who may show prevalence rates of over 40%."' Atopy is associated with IgE antibody to animal allergens and with asthma caused by animals but less closely with nasal, eye, or skin symptoms when these are present without chest symptoms."1013 To date, no relation between smoking and LAA has been identified. The survey oflaboratory workers described in a companion paper found a slight excess of current smokers in those with symptoms suggestive of occupational asthma.'0 We therefore examined data from this and two other populations to see if LAA is associated with smoking. Accepted 12 October 1987 Methods We have carried out cross sectional surveys of three populations occupationally exposed to animals. Survey A was ofpharmaceutical research workers and is described in a companion paper.'0 Survey B was of research workers at a tobacco company.'4 The laboratory had been open only a short time and all but three workers who had ever worked there were seen and the original survey record forms were available. Survey C was of pharmaceutical research workers. A study by company staff has been published" and our survey gave similar results.'5 The original forms were no longer available but some data were accessible on computer tape. The self completed questionnaire in survey A has been described and LAA symptoms were defined as chest, nose, eye, or skin symptoms provoked by rats, mice, guinea pigs, or rabbits or which were temporally related to occupational exposure to animals, improving with reductions in exposure.'0 Survey B used a similar definition in an interviewer administered questionnaire but included only rats, because only they were kept at the laboratory. An interviewer administered questionnaire was used at survey C and, apart from covering mice, guinea pigs, and rabbits as well as rats, was similar to that for survey B. The precise definition of LAA symptoms used to create the computer file on survey C was not clear so no data on symptoms are presented. At surveys A and C, smoking history was noted by 667 group.bmj.com on October 28, 2017 Published by http://oem.bmj.com/ Downloaded from